Standard Setting and Quality Regulation in Forensic Science

Standard setting and quality regulation in forensic science.

GeneWatch UK submission to the Home Office Consultation[1].

October 2006.

GeneWatch UK is a not-for-profit policy research group concerned with the science, ethics, policy and regulation of genetic technologies. Our aim is to ensure that genetics is used in the public interest. We welcome the opportunity to input to this consultation.

Our response is concerned with the use of DNA in forensic investigations and oversight of the National DNA Database (NDNAD).

We welcome the recognition in the consultation of a ‘regulatory gap’ and agree that a new regulator is needed. The aims of any new regulatory structure should be:

• to evaluate independently the quality, validity, cost-effectiveness and limitations of forensic services that may be procured by the police;
• to identify and address any broader ethical or social issues that might impact on public trust in forensic services;
• to operate in a transparent way, that is open to democratic scrutiny, and to consult and engage the public in important ethical decisions.

GeneWatch UK recognises that the FSS has played an important role in quality control of the National DNA Database (NDNAD). However, a number of issues of conflict of interest and lack of independent oversight have arisen, involving both the FSS and other suppliers. For example, the McFarland Review of the FSS found that “the overwhelming view of those with whom the issues were discussed was that the present custodian arrangements needed to be made more independent, more transparent and more accountable”.[2]

The main areas of concern that need to be addressed and are discussed below, are the use of new DNA techniques and oversight of the National DNA Database. Central to all these concerns is the need for public confidence that the police use of DNA is under adequate control. This requires far greater transparency and accountability. Many people – including children and victims of crime – are currently unsure whether their DNA has been retained or not, and equally unclear who may access their genetic samples and data and how they may be used. Many decisions – for example, regarding research uses of the Database – appear to have been made on a secretive and ad hoc basis, without proper scrutiny or oversight and with no public information, let alone engagement in the decisions that are made.

In addition, the decision to retain permanently all DNA taken on arrest for any recordable offence is undermining public trust in the police use of DNA. Although improved regulation is important, time limits on the retention of DNA profiles on the Database (related to the seriousness of the offence and whether or not an individual is convicted) are an additional important safeguard against misuse, as is the destruction of all individuals’ DNA samples, once DNA profiles have been successfully obtained from them. The aim of restoring public trust is therefore unlikely to be achieved until new legislation to address these issues is also adopted.

The use of new DNA techniques

A wide range of new, and relatively untested, DNA techniques are now being used. For example:

• The increasing use of Low Copy Number (LCN) DNA analysis – which allows a DNA profile to be extracted from a single cell – has led the Director of the Forensic Institute in Edinburgh to warn that innocent people may be wrongly identified as suspects as a consequence of being on the NDNAD.[3] LCN analysis and other new techniques such as “DNABoost”[4], increase the sensitivity of DNA analysis (allowing very small samples or mixed samples to be analysed, respectively) but also increase the chance of a false match between a scene of crime DNA sample and an individual’s DNA profile. These difficulties may be exacerbated when these techniques are combined.[5]

• An increasing number of services are available which aim to provide a “genetic photo-fit” of a suspect from a scene-of-crime DNA samples, predicting eye and hair colour and ethnic ancestry, for example. All these techniques have serious scientific limitations, but it is unclear whether police forces fully appreciate this.[6] In addition, ethical and social considerations may arise. For example, the use of ancestry testing during an investigation – as well as potentially giving misleading results[7] – may imply that a particular ethnic group, rather than an individual, is responsible for a crime. Misuse or misunderstanding of these techniques may mislead the police and lose public trust. The targeting of men of Caribbean origin (including police officers) for DNA samples by Operation Minstead (a hunt for a serial rapist in South London) has been particularly controversial.[8]

• The use of Single Nucleotide Polymorphisms (SNPs) as a new means of DNA identification is being strongly advocated by some commercial companies. SNP profiling is undoubtedly useful in identifying victims from degraded DNA samples following major disasters and acts of terrorism. However, its routine use in criminal investigations, or on the NDNAD, raises issues about statistical power, costs and privacy (since more genetic information may be revealed than when the standard system of STR profiling is used).[9] All these issues should be properly addressed before such techniques are used.

• New methods are being developed and adopted to attempt to collect DNA from a wider range of crime scenes (for example, ‘spit kits’ handed out to bus drivers and train staff); to speed up DNA analysis at crime scenes (via a mobile ‘lab in a van’), or to check the DNA, photographs, or fingerprints of individuals on the spot (for example, the new Criminal History System will reportedly allow officers will be able to input new information and access existing records from their patrol car or a hand-held device[10]). All these processes raise new issues about quality assurance and/or privacy and security of data.

The assessment of all these new techniques requires consideration not only of the potential for miscarriages of justice, and whether or not (and if so, how) the information they provide may be used as evidence in court. It is also important to consider the implications of their use during an investigation, including: potential to mislead the police; the likelihood of false matches and arrests; ethical, privacy and social implications; and potential impacts on public trust (including impacts on BME groups).

To address these issues the new regulator must:

• Have the necessary skills, powers and funding to make an independent assessment of the statistical validity of any new technique (for example by examining commercial ancestry databases), as well as inspecting and ensuring laboratory quality assurance.
• Be able to inspect and assess the impact of new techniques as applied in practice (for example, any reduction in reliability associated with using mobile laboratories and spit kits);
• Be able to identify and address any relevant social and ethical issues, so that public trust in forensic DNA technologies is not undermined;
• Assess cost-effectiveness, as a means to help police services make informed judgments about the services they may procure.

Whilst some issues may be considered during the suggested process of ‘type approval’, others (particularly statistical power and quality control) will depend on the details of the particular test on offer. For example, ancestry prediction may involve at least three different techniques (mitochondrial DNA, Y-chromosome DNA or bio-geographical tests) and the validity of the results will depend on the genetic differences and populations included in the databases of the particular company marketing the test (for example, the test marketed by US firm DNAPrint Genomics frequently makes surprising – and almost certainly misleading - claims about Native American ancestry[11],[12]).

Operation and oversight of the National DNA Database (NDNAD)

A number of important issues have arisen relating to access and uses of the NDNAD.

Familial searching of the NDNAD

Familial searching is sometimes used when a DNA profile from a crime scene does not match an individual’s profile on the Database. Since it is possible that a relative of the suspect is on the Database, looking for a partial match between profiles might identify a parent, child, brother or sister of the suspect, who can then be interviewed by the police. In 2004, the FSS reported that approximately 20 familial searches had been conducted and that a quarter of these had yielded ‘useful intelligence information’.

Familial searching usually produces a long list of names of people to be interviewed and raises ethical concerns because it is possible that it could reveal cases of paternity or non-paternity that the people interviewed did not know about, and also reveal to relatives who is on the Database. However, several high-profile cases have been solved using familial searching. Concerns about the ethical implications of this technique led to discussions between ACPO, the Home Office, the Information Commissioner and representatives of the Human Genetics Commission to agree a Memorandum of Understanding on its use.[13] However, this document is still not available for public scrutiny. The House of Commons Science and Technology Committee expressed concern that familial searching had been introduced in the absence of any Parliamentary debate about the merits of the approach and its ethical implications.[14] Any new system of regulation should require the ethical implications of new techniques such as familial searching to be widely debated and discussed before their introduction, and any rules about their use should be made public.

Overseas access

All 519 requests for details from the Database made by foreign countries since 2004 have been granted.[15] Details about why the requests were made are not publicly available, nor are the number of earlier requests. The European Union has proposed a Draft Framework Decision on exchanging information between law enforcement agencies in member states.[16] The British Government has argued that this should include direct on-line access to the National DNA Database by other countries. MPs have expressed concerns about giving widespread access to sensitive data on so many people – for example, many more people are on the UK database compared to France (which keeps DNA profiles only from people convicted of serious offences).[17] There is no formal, global provision to safeguard the confidentiality of DNA profiles when they are exchanged between police forces.[18] Any new regulator will have to consider the implications for public trust of this lack of transparency about decision-making and lack of international legal guarantees of confidentiality.

Research uses

Freedom of Information requests by GeneWatch UK to the NDNAD Board have shown that since the year 2000, 19 research projects have been allowed and 14 refused.[19] The requests revealed that stored DNA samples have been used for genetic studies of the male Y-chromosome, without the consent of the people involved, as part of a controversial attempt to predict ethnicity from DNA. This type of research could also inadvertently reveal other genetic characteristics such as a man’s risk of infertility.[20],[21] Despite numerous requests for information, the list of research projects is still incomplete and, in addition, the decision making process remains inadequate and unclear. Use of the NDNAD and associated samples without consent, for research unrelated to quality assurance of the database, is of questionable legality and risks a major loss of public trust. Many people on the Database would be unlikely to consent to genetic research on race, ancestry or ethnic appearance. Failure to involve people who are on the database in these decisions – whether they are prisoners or not – runs contrary to well-established ethical principles, intended to ensure that the vulnerable are not exploited.

Other search techniques

Speculative searching for a match between a DNA profile from a scene of crime (SOC sample) and an individual’s DNA profile (CJ sample) is now routine. However, there is some evidence that the NDNAD is sometimes used to search for the DNA profiles of named individuals.19 During research, other systems of searching have also been used (including searches by ethnic appearance category, or ‘having African name’, ‘having typical Muslim names’, or ‘having typical Hindu/Sikh names’).19 Searching for named individuals, or by category, raises new privacy issues because officers may obtain access to an individual’s (or group of people’s) DNA profile(s) and other private information, without any link to a crime scene having been made. There appears to be no ethical oversight over the potential use of such searches.

Sample and data retention by commercial suppliers

Records on the NDNAD may be linked back to the original DNA samples, using the barcode reference number. The samples are stored by the companies that analyse them, for an annual fee (70p per sample per year). They contain potentially unlimited genetic information (including, for example, some health-related information). Access to the samples therefore raises additional privacy concerns. Because of these concerns, the Human Genetics Commission has argued that samples should be destroyed once the DNA profiles used for identification purposes have been obtained.[22] Samples are destroyed in some other countries, such as Germany, and sample destruction is also part of Italy’s new law.[23] The cost-effectiveness and privacy implications of retaining samples permanently have never been independently assessed.

In addition, emails supplied to GeneWatch following a Freedom of Information request show that the commercial company LGC, which analyses some DNA samples for the police, has retained its own “mini-database” of DNA records (DNA profiles plus demographic data)[24]. The fact that commercial companies can keep copies of parts of the Database makes a mockery of claims that access to it is carefully restricted and controlled. Retention of data and samples by commercial suppliers are likely to have serious implications for public confidence in how the Database is managed.

Removal of records

Former Home Office Minister, Andy Burnham MP (in response to a PQ on 19th December 2005) stated that: "The decision whether to retain or remove a sample is an operational one for the chief constable of the police force which took it." Essentially the law allows for an individual's DNA profile to be added to the database, on arrest for any recordable offence, but does not make it compulsory to do so, or to keep their records and samples permanently. However, on 24th April 2006 the Association of Chief Police Officers (ACPO) produced a policy document giving guidance to Chief Constables on this matter. This 'suggested' that removal should only take place in exceptional circumstances and that Chief Constables may need support in making decisions on 'exceptional cases'.[25] ACPO have also provided Chief Constables with two sample letters, to deal with requests for removal. This process replaces the previous time limits for removal from the database (based on type of offence and whether a person had been convicted or not) with an unfair and arbitrary system, in which some unconvicted people are presumed more innocent than others. Whilst this situation can only properly be dealt with by a change in the law, any new regulator may have to deal with complaints about this process and/or make assessments of its fairness and cost-effectiveness.

The use of irrevocable consent for volunteers (including victims) has also been widely criticised14,22, particularly as they are not informed that, once entered on the Database, their DNA profiles and samples may be used for genetic research without their consent. Such uses are unethical and of questionable legality.

Lack of assessment of benefits

There were 20,349 direct DNAdetections in 2005-06 (i.e. there were 20,349 crimes detected in which a DNAmatch report was available)[26] – this compares to 20,489 in 2003/04 and 19,873 in 2004/05. This means that despite the ballooning of the Database to contain 1.5 million more individuals, the detection of crimes using DNA has not increased. The most likely explanation is that the people added to the Database in recent years (many of whom have never even been charged with any offence) are not at high risk of committing future crimes, or at least not the kind of crimes for which DNA evidence is relevant.[27] The success of the Database is in fact driven largely by the number of DNA profiles collected from crime scenes, not by the large number of individuals now being retained on the Database.[28] An important part of maintaining public trust would be an independent assessment of the impact of the expansion of the Database (particularly the permanent retention of DNA profiles and samples from a large number of innocent individuals) on the detection of crime, and on successful prosecutions, including its cost-effectiveness.

To address these issues the new regulator must be:

• independent of the Home Office and forensic service providers;
• responsive to public concerns about uses of the Database – including the concerns of people on the Database, many of whom may never have been convicted of any offence (including victims of an offence), or who may be vulnerable to exploitation (such as mentally ill offenders and children, and members of BME groups);
• transparent and consult widely, allowing public input to decision-making, including from Black and Ethnic Minority Groups;
• able to take account of ethical concerns and weigh up the privacy and rights implications of new uses, including retention policies, against any benefits in tackling crime.

Regulatory proposals